Created
November 25, 2021 08:35
-
-
Save sordina/b3df5f9bd0d6adac2a40dcdb04a649ab to your computer and use it in GitHub Desktop.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* Les: Getting more comfortable with HoTT - Talk: "Bluffers Guide to HoTT" | |
History - Barber's Paradoxes, etc. | |
Russell stratifying set theory | |
Martin Löf in the 1970s | |
In TT - Say term x is of type t -> `x : t` | |
This is a judgement | |
Curry howard - Types and terms = Conjecture/proposition and Proof | |
Things become interesting in the presence of equality | |
equality is always with respect to some type | |
exists t -> x:t = y:t | |
can also think of this as a type: The type of all proofs that x=y | |
e.g. 2+3 = 3+2 | |
Proof by reduciton | |
Proof by commutivity of addition | |
In TT you can have different proofs of things being equal, | |
or think of things being equal in two different ways | |
Proofs are also terms, so you can discuss equality of proofs | |
There is a structure where you can have different proofs of a proposition, | |
but those proofs may or may not be "equal" | |
topology or homotopy analogies come into play here | |
Equality proofs are paths in the space of objects | |
There can also be paths between paths equating proofs | |
There is an infinite hierarchy of these correspondence proofs | |
Ties into Vladimir Vodovosky's work - Univalence Axiom | |
Isomorphism and equality aren't always the same thing | |
Mathematicians sometimes will after proof of isomorphism treat objects as equal, | |
and lose the provenance | |
"Equivalence is equivalent to equality" - Formalised in TT | |
This is useful because it provides a fomal basis for common idioms in mathematics | |
HoTT people are promoting HoTT as a new foundation of Math | |
Vod: Provides a better foundation for mathematics and points out that it matches | |
the practice of mathematicians better | |
No types in ST | |
Bob-Harper: | |
"Doing math in ST is like programming | |
in machine code and doing it in TT is like using a HL language" | |
UniMath - Coq library that formalises a lot of math as code | |
HoTT people did the development of the book as a colab through git! | |
100s of mathematicians contributed | |
Two formulations of the reals | |
- Cauchy sequences | |
- Dedekind cuts | |
In traditional math these are two approaches to the same thing | |
Some work in HoTT was to define both of these in terms of TT and | |
shown to be inequivalent without LEM | |
i.e. decidable, vs. undecidable | |
Proof by contradiction | |
One of the problems with HOTT - when invoking univalent axiom you lose computability of proofs | |
Bob Harper developed similar approach - CCTT "Cartesian Cubical Type Theory" | |
A slightly different approach that preserves computation. | |
Back to LEM - In TT - Based on intuitionistic logic - Two things can happen | |
- LEM not baked in but can assume if convenient | |
- In many situations you can prove LEM in certain circumstances WO axiom | |
E.g. Nats | |
In FO Predicate calculus LEM baked in | |
Often told "in constructive logic you can't use proof by contradiction" | |
That's a characature of the actual situation - of course you can use it in TT - it's definitional | |
In constructive logic - while you can't prove p by assuming not p and arriving at contradition, | |
But you can prove not p by equivalent procedure |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment